[medved]

Hy


I was thinking about feasibility of making linear stepper by myself. For now, it is all "theory" and none praxis, but i would like to hear from others, do you think the theory is right? Or maybe someone has allready done it in praxis?

For building the motor, only two thin soft iron sheets should be machined on mill:
a)one for stator - where notches are equally spaced -lets say milled with one milimeter diameter mill

b)the other iron sheet for "rotor" -with 4 groups of notches, relatively displaced to each other for 1/4 of a phase.


All other components could be "hand made"-without need for precision machining. Isolated sheets of metal are stacked together to form a core, glued to stator and rotor sheet, and copper wire is hand wound on to metal stacks. Connect the wires, find some linear guides, and the motor is ready to go

If distance between two notches is 2mm then each step is 1/4 of that, 0.5mm. When using half step scheme then resolution is 0.25mm and with microstepping even better.

I added some pictures to help visualise the idea. The mysterious pipes on pictures represent copper windings. I hope i didnt make some fundamental electrotechnic mistakes.



What do you think, is this idea worth a try? What details have i overlook?

[joecnc2006]

How would you keep debri from entering it, it would have to be a solid sealed system.

[greybeard]

I fear the soft iron plate attached to the stacks with the windings on effectively "short circuits" the magnetic fields of the windings. So very little field strength will be available to induce a magnetic field in the lowest plate.
Thus there will be very little interaction between the two plates.
The individual coils need to be wound around stacks which have a magnetic circuit completed through a plate on top of them in your diagram.

[ESjaavik]

In your drawing the pole pieces looks like rods with one end up in the air. If you make them like horseshoes instead, that will close the magnetic flux. Then you will have a linear Variable reluctance stepper motor (no permanent magnets)

If you take a toroid transformer, discard the windings, split it in 2 to get 2 C-shaped parts. Then you mill teeth in the places where you cut it. Then put windings on each of them. Then you have your pole pieces (in theory).

Making it is quite another story though. It has been done in practice. You can buy it (if you have loads of money).

[medved]

Hi

I fear the soft iron plate attached to the stacks with the windings on effectively "short circuits" the magnetic fields of the windings.

If i understand right, you are refering to the upper plate-the "rotor". My picture was only shematic, in reallity 4 units would be much more apart. To prevent shorting of magnetic field, 4 separate iron sheets could be attached to non magnetic aluminum that would hold them apart. They would be glued first to aluminum, and AFTER that machined- this would enable precision without problem with proper alignment.

In your drawing the pole pieces looks like rods with one end up in the air. If you make them like horseshoes instead, that will close the magnetic flux.

My idea was to use "HALF-POLAR"(?) motor, so upper 4 windings would use only "botom"-"south" part of magnetic pole, and lower winding-stator would use only upper part of its magnetic field- its"north pole". Like two permanent magnets attach to each other as theirs north and south pole is brought together.


I got a little confused with Greybeard and ESjaavik's answer, but then i figured it out.
When Greybeard sais magnetic field would turn-around, close itself before it would arrive to lower stator - he is reffering to my "half-polar" principle. But when he and ESjaavik are proposing to make U-shape windings, they are reffering to different principle than mine - they reffer to principle of real linear steppers.
You are both is right, your idea is better, because it doesnt need the lower-stator to have any windings, it can be a plane piece of metal with notches machined in it!

I added pictures form Isel catalogue. I believe it shows the principle you are talking about. Dont let the permanent magnets on picture confuse you-they have nothing to do with stepping motion, they are there just to keep "rotor" attached to stator-because "rotor" is like hovercraft flying on layer of compressed air. But what is confusing me, is how the field is closed. On picture it looks like the flux goes in at one notch's hill and comes out at the valley of the neighbour notch - that is just milimeters apart - wouldn't these two fluxes interact and somehow aninhilate?
Besides only two units are hown on Isel picture, there need to be four - so i have trouble imagining how the stepping sequence works.

I have added my modified pictures, do you think this system would work? As i said, 8 separate iron sheets would be glued to nonmagnetic aluminum and AFTER that notches would be machined into iron - this way there would be no problems with alignment. Also a U-shaped stack of iron would have to be machined-but they would not have to be very precise. Dont you think that such design could be built by amateur?

Such self made units could be used in very big mills -for plasma cutting, or for cutting foam - I think it could be cheaper than buying rack and pinion. And it could be used in very small units - coupled with linear scales - for EDM or for ECG.


I looked at Isel technical data, such motors can produce from 1.5kg to 54kg of force and maintain precision of 2 mikrometers!!!

[greybeard]

Medved - Could you add some labels to your diagram - text on each part. This would be very helpful.
Could you also give a link to the isel page, then I might be able to get a translation of their text. Then I can better understand how your idea is different from theirs.

Simplification of a previous design may let you make a diy version, but be careful not to lose the physics principles by which the first one worked.

.... and keep trying.

[ESjaavik]

@Medved,
Look again at the Isel drawing and you will see that the flux of the permanent magnets and the coils cancels out in the piece that is "over the valleys". The flux lines are shown in blue with arrows indicating the direction. The two inner flux paths are caused by the coils, the remaining lines by the PM's.

And the stepping sequence is the one you recognize from a standard 2ph stepper. 2 coils with 2 possible directions of current (and flux) flow makes 4 possible energized states.

The arrangement shown makes it a "Permanenterregter Reluktanz-Schrittmotor" or as they would say further west: a hybrid step motor.

I assume you already did know your "Jones on Steppers", as that is mandatory reading before posting here?
http://www.cs.uiowa.edu/~jones/step/

So, off you go to build it. If it does'nt work as a stepper, turn it upside down, reposition the windings and use it as a magnetic table.

[greybeard]

ES - Having read your post and looked again at the isel diagram, I realise that I haven't the faintest idea how this thing is supposed to work. So I am retiring from this thread before I make a complete idiot of myself. (chair)

[ESjaavik]

Hope I didn't scare you.

The Isel drawing would have been much easier to understand if the 2 flux paths did not have the same colour. Look at the drawing again. Where the arrows through one leg point in opposing directions, they cancel out each other. So there is no magnetic force acting.Where they point the same direction, they enforce each other.

Regarding making an idiot of yourself: A person trying hard to learn never looks like an idiot to me!

[CJL5585]

MedVed:
This concept is used in the Japanese Magnetic Levitation Train system. There was a program on the Discovery Channel about the design.

You are a little late on a patent. Sorry.

Jerry

[MrWild]

"I assume you already did know your "Jones on Steppers", as that is mandatory reading before posting here?
http://www.cs.uiowa.edu/~jones/step/"



Thanks. This is a great place. I've been devouring Jones on Steppers and have a question regarding power.

If an L296 has a ceiling for amps and a Mosfet driver can control Mosfets to make an H bridge, could the hi amp H bridge be swapped in an PIC/L296 circuit? All the cautions about clamping diodes and capacitors would have to be met of course, but if there is PIC chip/code for current limting and microstepping that works with an L296, could the H bridges be swapped around? 6a. - 9a. per phase could be possible.

If this is impractical, why is it impractical? It seems this would have been done already and if it isn't around it got abandoned for some reason.

This is where I got this thought:

"While integrated H-bridges are not available for very high currents or very high voltages, there are well designed components on the market to simplify the construction of H-bridges from discrete switches. For example, International Rectifier sells a line of half H-bridge drivers; two of these chips plus 4 MOSFET switching transistors suffice to build an H-bridge. The IR2101, IR2102 and IR2103 are basic half H-bridge drivers. Each of these chips has 2 logic inputs to directly control the two switching transistors on one leg of an H-bridge. The IR2104 and IR2111 have similar output-side logic for controlling the switches of an H-bridge, but they also include input-side logic that, in some applications, may reduce the need for external logic. In particular, the 2104 includes an enable input, so that 4 2104 chips plus 8 switching transistors can replace an L293 with no need for additional logic.

The data sheet for the Microchip (formerly Telcom Semiconductor) TC4467 family of quad CMOS drivers includes information on how to use drivers in this family to drive the power MOSFETs of H-bridges running at up to 15 volts."
Jones, Douglas W. http://www.cs.uiowa.edu/~jones/step/circuits.html

And Microchip offers samples

[medved]

Now, this Isel diagram is like roller-coaster of comprehension to me: now i understand, the next moment i dont understand.

Where the arrows through one leg point in opposing directions, they cancel out each other. So there is no magnetic force acting.Where they point the same direction, they enforce each other.

Put this way, this diagram finally makes sense. Congratulations to ESjaavik for deciphering it. I was suspecting it was a bipolar motor, but the placement of notches didnt make any sense to me until you mentioned that some fields cancel each other with permanent mangnet and other fields enhance due to permanent magnet. The text in Isel descripton somehow mislead me, so i thought permanent magnet has nothing to with stepping sequences, but it has everything to do with it.


Looks like this "rotor" has to be super precision built, made of alternating stacks of permanent magnet and iron with windings (or can the permanent magnet simply be put on the top of the windings?).
Such design would certanly be out of reach for amateur builder.

But i was thinking of easyer design. UNIPOLAR motor, so there would be 4 separate windings, all precisions cutting of notches is done in one milling session. U-shaped windings are then glued on the top of the nothces and dont have to be precision built or precision positioned. The basic principle of operation would be the same as with Isel motor: flux from U-shaped electromagnet is closed through lower stator.



Greybeard, I copyed Isel pictures from Isel pdf catalogue, which is 95Mb big. There is their webpage in english http://www.iselautomation.net but i am affraid they dont have small .pdf on linear motors, only a complete (95Mb) "katalog_2004_komplett.pdf". I will attach text file, so you can translate it at some translation webpage.

[greybeard]

ES - thanks for your concern, and for pointing out the improvement that separate flux path colours would have given. I begin to see a little light. Could you improve matters still further by translating the word Laufer and winkelung. I guess the later is coil, but the former seems to point to the magnets(?). My student german dictionary is no help.
Like CJL5585, I assumed I was looking at a rehash of the maglev motor(bangsquiet drum for British engineer Eric Braithwaite)

[medved]

Could you improve matters still further by translating the word Laufer and winkelung. I guess the later is coil, but the former seems to point to the magnets(?).

Winkelung are windings i think, and Laufer would be "rotor", that is, it is a complete moving part UNIT(as opposed to static part-the stator), like cart.

[ESjaavik]

Could you improve matters still further by translating the word Laufer and winkelung. I guess the later is coil, but the former seems to point to the magnets(?).

Laufer = runner. In this case it's the complete moving assembly with magnets, coils and pole pieces.

Winkelung = angle. But in this case it is probably a misspelling for Wicklung=winding.

It is understandable your dictionary does not cover Winkelung. I guess it is not much used outside dog posture evaluation and some other odd situations.

[greybeard]

"dog posture evaluation" - mind boggling. Do lamposts come into it ?